It’s nearly a week since I posted this question and I’ve not had any responses, which I guess means nobody knows why this happens! So should I file a bug report?
I wasn’t aware that this topic existed, whoops.
This is the first time I encounter the issue, but it apparently was there for a long time indeed.
Last week we added new charting functionality, also enhancing the datapoint endpoint.
We changed the HTTP API spec to use a querying approach using POST and a body with details.
This is also to future-proof the endpoint for features we have planned.
So I guess that update would solve the problem since its spec got changed, and we’re not using DatapointInterval anymore. Updating the manager to the most recent version should do it.
Be aware of the migration steps if using a production server.
I put some comments down in the GitHub issue on the topic as well.